Segmentation Error or decrease the runtime

[Leronira01]

Hello everyone,

First, sorry for my English. I hope you understand.

See, I have a problem of Segmentation Error.
I already saw many post about that and I concluded that it was because of inestability of the model, so I change my timestep of 60 to 30 and it worked.
However, I seem the time of run is too much (about 6 hours for 30 hours). Is it normal? Is it a problem of my configuration? I will post my namelist wps and input too.
Any suggests to short the time? Please I need the time decrease without impact in the confiability of the results. (Should I change the physics?)



Thanks!

 

[Ming Chen]

This is a triply-neted case with the smallest grid interval of 1km. I guess it does take time to run this case, although the wall clock time is also dependent on how many processors you use.

 

[Leronira01]

This is a triply-neted case with the smallest grid interval of 1km. I guess it does take time to run this case, although the wall clock time is also dependent on how many processors you use.

Thanks for the response Mr. Chen!
I dont'know if the term is correct but I'm using 8 cores.

 

[Ming Chen]

Do you mean 8 processors? If so, then 16 processors might be better for this case.
If you mean 8 cores, then how many processors does each core contain?